
1 minute read
International Journal for Research in Applied Science & Engineering Technology (IJRASET)
from Predicting Stability of Slope by Amplitude and Coherence Using a Naïve Bayes Classifier
by IJRASET

ISSN: 2321-9653; IC Value: 45.98; SJ Impact Factor: 7.538
Advertisement
Volume 11 Issue III Mar 2023- Available at www.ijraset.com
The principal objective of this study is to provide the range for coherence and amplitude to help determine the stability of slopes with high confidence regardless of slope deformation and inverse velocity values. As mentioned earlier, the GNB assumes that the contributions are independent. To determine if coherence and amplitude are independent, a heat map, shown in Fig 4, was constructed by using the Pearson correlation coefficient score which measures the linear relationship strength between two variables. The heat map reveals that both coherence and amplitude have a correlation coefficient between 0 to -0.2, which means they are not completely independent of each other. Despite the moderate dependency of the two attributes, we relied on many studies such as Domingos et al. (Domingos and Pazzani 1997) which showed that a Bayesian classifier still can perform optimally even if the independence assumption for contributions is not fully established. Hence, wecontinued using theNBC for this study.



